ELEMENTS IN IGNEOUS ROCKS—WASHINGTON. 283 
rarity, it being understood that the maxima are seldom attained and 
that very frequently these elements are wholly absent. 
Zirconium is much less common than the chemically related 
titanium, and seldom exceeds 0.20 per cent, though in some very ex- 
ceptional cases it may reach 2 per cent. Chromium seldom occurs 
in amounts above 0.5 per cent, though a few rocks are known in which 
it is reported to range between 2 and 3 per cent. Nickel seldom ex- 
ceeds 0.20 per cent, and the allied cobalt is scarcely ever present in 
more than mere traces. The maximum amount of copper found in 
unaltered igneous rocks may be placed at about 0.2 per cent of CuO. 
Barium almost always exceeds strontium in quantity, but only very 
exceptionally exceeds 0.25 per cent, though some rocks are known in 
which about 1 per cent is present; while the amount of strontium is 
usually much less than 0.1, but may occasionally reach 0.3 per cent 
in the rocks very high in barium. Although figures of 0.1 or 0.2 
per cent have been reported for lithium, these are somewhat doubtful, 
and it seldom occurs in more than spectroscopic traces. Sulphur and 
chlorine may both be present up to 2 per cent or slightly more, but 
both usually occur only in tenths of a per cent, and the latter amount 
is the maximum for fluorine. Of the other minor constituents the 
amounts present are so small as usually to be insignificant except as 
to their actual presence. 
Ili. THE AVERAGE COMPOSITION, 
The estimation of the average composition of igneous rocks as a 
whole is not such a simple matter as may be thought at first, because 
of several complicating factors which should be taken into account, 
and certain corrections in some of our data which should be made, to 
obtain fairly satisfactory results. We are not yet in a position to 
make precise allowance for these, into the discussion of which we can 
not enter here, so that the results so far obtained can be regarded as 
but first approximations, only roughly correct, but of some value. 
The most recent and reliable calculations are three made suc- 
cessively by Prof. F. W. Clarke, of the United States Geological Sur- 
vey, two of the igneous rocks of the British Isles by Prof. A. Harker, 
of Cambridge, and one made by myself some years ago.” 
Clarke’s latest estimate is based on more than a thousand analyses 
of igneous rocks of the United States made by the chemists of the 
Geological Survey, and mine is based on about one thousand eight 
a}, W. Clarke, Bulletin of the U. S. Geological Survey, No. 148, p. 12 (1897) ; 
No. 168, p. 14 (1900) ; No. 228, p. 18 (1904); A. Harker, Geological Magazine, 
vol. 36, p. 18 (1899) ; Tert. Ign. Rocks of Skye, p. 416 (1904) ; H. S. Washing- 
ton, Professional Paper No. 14, U. S. Geological Survey, p. 108 (1903). I am 
at present engaged in new calculations of the average rock, based on more than 
three thousand analyses, but this is not yet ready for publication. 
